Search

A fast back-projection (FBP) imaging algorithm is proposed to generate 3-D uniformly balanced images of an object embedded in a layered structure based on a combined near-field monostatic and bistatic synthetic aperture radar (SAR) system. The spectral layered Green’s function is used in the FBP algorithm. Like conventional real-time imaging algorithms, the complexity of the imaging function is reduced via the stationary phase method, and the fast Fourier transform is implemented to accelerate the computation. Unlike conventional real-time imaging algorithms, the singularity problem associated with Green’s function and antenna pattern is solved in this proposed algorithm by using a point spread function (PSF). For a real-time imaging system, besides fast image generation, an efficient data acquisition is needed. A combined monostatic and bistatic SAR configuration is used here to decrease the data collection time. The performance of this proposed FBP method is demonstrated by measurements of a stratified environment. [ABSTRACT FROM AUTHOR]

We present a systematic method for designing dual-polarized, platform-mounted, high-frequency (HF) antennas for near vertical incidence skywave (NVIS) applications. To overcome the bandwidth (BW) and efficiency limitations of electrically small antennas, the characteristic mode (CM) theory is used to exploit the metallic platform as the main radiator. Low-profile coupling elements are mounted on different locations of the platform to excite two orthogonal, horizontally polarized CMs. This way, dual-polarized operation is achieved with a high isolation and sufficiently wide BW. In the design process, several key practical issues are taken into account and their impacts on the performance of the antenna are examined carefully. These include the type and design of the coupling elements, the presence and type of the earth, and the non-idealities and power handling capability in the impedance matching network components. The simulation results of both a full-scale and a 1:50 scaled model of the proposed structure are presented. The scaled-model prototype was fabricated and characterized in the presence of a ground plane emulating the properties of real earth at the scaled frequency of operation. The measurement results show a good agreement with the simulations, demonstrating the efficacy of the proposed approach in designing dual-polarized, platform-based HF antennas. [ABSTRACT FROM AUTHOR]

Estimation of wall parameters is critical for focused through-the-wall imaging (TWI). In this paper, a novel algorithm based on pole extraction and wall reflection coefficients is presented to accurately estimate the essential parameters of a multi-layered wall. The generalized pencil of function (GPOF) is applied to decompose the frequency-domain received signal into a summation of ordered exponential terms and phase delays. Matching the terms with the multiple wall reflections provides equations for extracting the parameters. A frequency-domain calibration process is proposed to remove the antenna response from the measured wall response. The formulation of the proposed parameter extraction and calibration algorithms is presented along with the simulated and measured results for validation. The efficacy of the method is demonstrated on a real TWI system. [ABSTRACT FROM AUTHOR]

The uniform diffraction tomographic (UDT) algorithm corrects discontinuities associated with conventional DT algorithms when applied to imaging of objects buried in layered media. The cause of the discontinuities in DT is in the implementation of the stationary phase solution for evaluating the spectral integral. The usual approach is to make a far-field approximation assuming that the depth is large. However, this is not appropriate for near-field imaging. UDT overcomes this issue by choosing the correct large asymptotic parameter in the exponent of the spectral integrand. In this paper, the original 2-D scalar UDT algorithm is extended to a fully vector 3-D solution involving the dyadic Green’s function for the layered media, the polarization of the gain pattern of the probe antenna, and the stationary phase evaluation of a 2-D spectral integral. The 3-D UDT is demonstrated to provide smoothly continuous images, and like the conventional DT, the fast Fourier transform is implemented to generate images in real time. 3-D numerical simulations and experimental results are presented to verify the accuracy and efficacy of the 3-D UDT algorithm. [ABSTRACT FROM AUTHOR]

A 2-D near-field uniform diffraction tomographic (UDT) imaging algorithm is formulated to generate images of targets embedded in a layered structure. The conventional diffraction tomography (DT) improperly applies the stationary phase method for stratified environments to reduce the innermost spectral integral. In DT, the large argument is assumed to be the depth, which is not appropriate for near-field imaging. This results in amplitude discontinuities occurring at the interfaces between adjacent layers. The correct large argument is the free space wavenumber as used in high-frequency asymptotic solutions. The UDT therefore yields uniformly continuous images across the interfaces. And like the DT, the UDT retains the fast Fourier transform relation in the algorithm for generating images very efficiently. Numerical and experimental image comparisons between DT and UDT for objects buried in stratified environments are presented to demonstrate the efficacy of the proposed UDT method. [ABSTRACT FROM PUBLISHER]